Background: Stratifying patients with paroxysmal or short-term persistent atrial fibrillation (AF) who are at greater risk of developing permanent AF is challenging. There are studies on predicting persistent AF in patients with paroxysmal AF, but studies evaluating natural course of progression to permanent AF are rare. (1, 2) Aim of our prospective study was to evaluate utility of routine demographic, clinical, laboratory and echocardiography parameters, together with evaluated risk scores in prediction of AF progression to a permanent form.

Methods: In the period of 30 months we prospectively recruited 409 patients with paroxysmal or short-term persistent AF who were treated at discretion of the referral cardiologist in University Hospital Dubrava and followed them for a median follow-up time of 21 months. Clinical, laboratory, and routine echocardiographic parameters were collected. Endpoint was progression to permanent AF when further attempts to restore sinus rhythm were abandoned.

Results: Out of 409 patients with non-permanent AF, 109 (26.6%) progressed to permanent AF during follow up. Patients who progressed had significantly lower estimated glomerular filtration rate (eGFR), higher age, body mass index, CHA₂DS₂-VASc score, HATCH score, LADS score, LA diameter, C-reactive protein, red cell distribution width (RDW) and mean platelet volume (MPV) levels, and also higher proportions of arterial hypertension and previous stroke. In multivariate Cox regression model only increased left atrium (LA) diameter, and increased RDW showed significant independent association with progression. When corrected for LA size at 45 mm and RDW level at 14.5% LADS score dichotomized at <4 did not show significant effect on progression, whereas corrected HATCH score dichotomized at <3 did show significant independent effect on AF progression during follow up and had the best negative predictive value of 0.87. None of the observed parameters showed a positive predictive value for AF progression >0.60.

Conclusion: LA size and RDW levels strongly moderate estimated risk of AF progression. Although it is still challenging to predict progression, patients with LA size ≤45 mm and RDW level ≤14.5% and a HATCH score <3 had the least probability of AF progression, and are most probably the best candidates for rhythm control strategies.